Metraipolis › archive › dc_0eb8407f8dac

Tight Stochastic Condition-Number Dependence in Nonconvex-Strongly-Concave Minimax Optimization — arXiv:2609.30877

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.30877
License (as archived)
CC0-1.0
Archived
2026-10-08T02:19:13.838Z
SHA-256 of the archived text
558eb6e8bb82893fa75f639bf8d118acd55b4b3bf3fcf8f30608896a1420ed10
Stable URL of this page
https://metraipolis.com/documento/dc_0eb8407f8dac

How to cite

Tight Stochastic Condition-Number Dependence in Nonconvex-Strongly-Concave Minimax Optimization — arXiv:2609.30877 — arXiv (en), https://arxiv.org/abs/2609.30877 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0eb8407f8dac (Metraipolis archive dc_0eb8407f8dac, sha256 558eb6e8bb82893fa75f639bf8d118acd55b4b3bf3fcf8f30608896a1420ed10)
BibTeX
@misc{dc0eb8407f8dac,
  title = {Tight Stochastic Condition-Number Dependence in Nonconvex-Strongly-Concave Minimax Optimization — arXiv:2609.30877},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.30877},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0eb8407f8dac; sha256 558eb6e8bb82893fa75f639bf8d118acd55b4b3bf3fcf8f30608896a1420ed10},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.30877. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We study whether the linear condition-number dependence in the stochastic complexity of SAPD+ is necessary for nonconvex-strongly-concave minimax optimization. For jointly $L$-smooth objectives with dual strong-concavity parameter $\mu$, we prove a lower bound that matches the SAPD+ upper bound under the same Moreau-envelope stationarity criterion and the same primal-dual initialization gap. Specifically, when $\sigma\ge\varepsilon$, the worst-case complexity of zero-respecting algorithms is $\Theta(\kappa LG\sigma^2\varepsilon^{-4})$ in the stated accuracy regime, where $\kappa=L/\mu$, $G$ bounds the initial primal-dual gap, and $\sigma^2$ bounds the variance of a general unbiased first-order oracle. The lower bound is realized on a smooth problem class with a bounded dual box. Our construction routes each link of a nonconvex zero-chain through a dual gradient of magnitude proportional to $\varepsilon/\sqrt{\kappa}$, while an undiscovered primal coordinate prevents stationarity. It also yields the primal-gradient lower bound $\Omega(L\Delta(\sqrt{\kappa}\varepsilon^{-2}+\kappa\sigma^2\varepsilon^{-4}))$ after combination with the known deterministic bound, where $\Delta$ bounds the initial primal function gap.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-28 in cs.LG).
Categorie arXiv: Optimization and Control (math.OC) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.30877 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-08T01:43:06.799Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-28?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).